NMR and X-ray structural studies on 3-benzyl-8-bromoadenine

Article abstract of DOI:10.1002/jhet.733

8-Bromoadenine was benzylated in the presence of base to give a mixture of two regioisomers. One was easily recognized as 9-benzyl-8-bromoadenine, but the other structure could not be determined with absolute certainty by NMR. Therefore, X-ray crystallogr

Full text of DOI:10.1002/jhet.733

November 2011
NMR and X-ray Structural Studies on 3-Benzyl-8-bromoadenine
1375
¨
Huey-San Melanie Siah, Carl Henrik Gorbitz, and Lise-Lotte Gundersen*
Department of Chemistry, University of Oslo, Blindern, N-0315 Oslo, Norway
*E-mail: l.l.gundersen@kjemi.uio.no
Received August 9, 2010
DOI 10.1002/jhet.733
Published online 30 June 2011 in Wiley Online Library (wileyonlinelibrary.com).
8-Bromoadenine was benzylated in the presence of base to give a mixture of two regioisomers. One was
easily recognized as 9-benzyl-8-bromoadenine, but the other structure could not be determined with absolute
certainty by NMR. Therefore, X-ray crystallography was used to prove that the benzyl group was attached to
N-3. Furthermore, it is shown that the 3-benzyl adenine derivative exists as the amine tautomer both in the
crystalline state as well as in solution (DMSO-d6), with restricted rotation around the N6AC6 bond.
J. Heterocyclic Chem., 48, 1375 (2011).
INTRODUCTION
temperatures and the purification of products was rather
tedious. The spectral data for our major product were in
good agreement with what has been reported for com-
pound 3 before. However, despite the fact previous litera-
ture structure elucidation is claimed to be based on
1H-13C HMQC and HMBC NMR (no details given) [4],
we found our HMBC data to be inconclusive. The CH2
protons correlate to two carbon shifts; the C-2 at 143.9
ppm and a peak at 149.8 ppm (quaternary C). Unfortu-
nately, it was not possible to determine if the latter peak
was the C-4 or C-6 shift, and hence, it was not possible
to establish if the benzyl group was situated at N-3 or N-
1. Both the peak at 149.9 ppm and a second peak (quater-
nary C) at 153.6 ppm correlated to H-2 and neither of
them correlated with the NH2 in the HMBC spectrum.
Because there are several reports on the formation of N-
3 alkylated 8-bromoadenine, where the structure elucida-
Generally, reaction of adenine with alkyl halides under
neutral conditions gives the 3-alkylated product as the
major isomer. However, selective N-9 alkylation is often
seen when the reaction is performed in the presence of a
base [1]. The numbering of the adenine system is shown
in Figure 1. When 8-bromoadenine is alkylated under ba-
sic conditions, the selectivity for the 9-position is gener-
ally reduced, and mixtures of the 3- and 9-alkylated iso-
mers are often reported [2–6]. The isomeric distribution
is qualitatively the same in Cu-mediated N-arylation of
8-bromoadenine [7]. When 8-bromoadenine was benzy-
lated in our laboratories under basic conditions, we also
obtained two isomers, but we found that the position of
the benzyl group in one of the isomers could not be
determined with absolute certainty by NMR spectros-
1
copy. Furthermore, the H-NMR spectrum of the same
1
tion is reported to be based on H-13C HMQC and HMBC
compound showed some intriguing features (discussed
below) and, as a result, we performed thorough NMR
and X-ray structural studies on this compound.
NMR [4,5,7], which in fact may not be conclusive, we
decided to determine with absolute certainty whether the
major product formed was the isomer 3 or 4. For these pur-
poses, we turned to X-ray crystallography. Before this
investigation, crystal structures of ten 8-bromoadenines
were available in the Cambridge Structural Database
(CSD, Version 5.31 of November 2009 [8]). All these mol-
ecules carry an additional ethyl group or a sugar moiety,
which is always attached to N-9 as for isomer 2.
The result of the X-ray structural investigation shown
in Figure 1 confirms that isomer 3 has been crystallized.
This is not only the first ever X-ray structure of an N-3
functionalized 8-bromoadenine but in fact also the first
N-3 functionalized adenine where neither N-7 nor N-9
act as ligands for metal ions or carry additional
RESULTS AND DISCUSSION
8-Bromoadenine (1) [2] was reacted with benzyl bro-
mide in the presence of K2CO2 at ambient temperature
as shown in Scheme 1. The minor product was easily
identified as the 9-benzylated isomer 2 based on com-
parison with data reported before [4] as well as H-13C
1
HMQC and HMBC NMR experiments. Compound 2 is
reported to be the major isomer when the benzylation
was performed at higher temperature [4], but in our
hands more than two products were formed at elevated
C
V
2011 HeteroCorporation

¨
H.-S. M. Siah, C. H. Gorbitz, and L.-L. Gundersen
1376
Vol 48
Figure 1. The two adenine molecules in the asymmetric unit of 3 with atomic numbering indicated (not in correct crystallographic positions rela-
tive to each other). Displacement ellipsoids are drawn at the 50% probability level, H atoms are spheres of arbitrary size. Important bond lengths
˚
(in A) have been indicated, estimated standard deviations are 0.005–0.006 A. The different orientations of the benzyl groups are defined by the tor-
˚
sion angles C2AAN3AAC10AAC11A
N3BAC10BAC11BAC12B ¼ 177.7(5)ꢁ.
¼
70.6(6)ꢁ, N3AAC10AAC11AAC12A
¼
76.7(6)ꢁ, C2BAN3BAC10BAC11B
¼
90.8(6)ꢁ, and
functional groups. Bond lengths and bond angles are
roughly the same for the two molecules in the asymmet-
ric unit and correspond to the regular amino tautomer.
Both amino H-atoms were easily located in the electron
density map and participate in strong hydrogen bonds as
shown in Figure 2.
(Fig. 4). Computational calculations have indicated that
the amino tautomer of 3-methyladenine is more stable
than any of the possible imine forms in the gas phase [9],
and these calculations support the hypothesis shown in
Figure 4.
1H-15N HSQC NMR spectroscopy showed that both
NH signals correlated to the same nitrogen signal at
125.7 ppm (relative to 15NH3) proving that compound
3 exists as the amino tautomer 3a in solution (DMSO-
d6). It is interesting to note that the NH2 in compound
3 is shifted substantially downfield compared with for
instance the NH2 in 9-methyladenine which resonances
at 79.6 ppm (ꢀ300.9 ppm relative to Me15NO2) [10]
indicating more sp2 character for the N6 in compound
3a. The coalescence temperature for the NHa and
NHb was not determined, but at 65ꢁC only one broad
signal from the NH2 could be seen in the 1H-NMR
spectrum.
1
In the H-NMR spectrum of compound 3 in DMSO-d6
solution, there are two distinct NH signals (8.07 and 8.23
ppm), in contrast to the spectrum of the 9-benzylated
product 2, where one broad singlet for the NH2-group is
seen. Two NH signals have also been observed in
spectra of other 3-alkylated adenines [3–6], but this phe-
nomenon has only been discussed once previously and
the hypothesis presented was that the compound studied
must exist in solution as an imine tautomer. However,
no experimental evidence is given [6]. We thought that
the reason for the splitting of the NH signals could either
be that compound 3 exists as an imine tautomer in solu-
tion 3b or 3c (Fig. 3) or that there is restricted rotation
around the N6AC6 bond leading to two resonances for
the NHa and NHb. This could be the case if resonance
forms 3a0 and or 3a00 contribute to the structure of 3a
In summary, we have shown with absolute certainty
that benzylation of 8-bromoadenine in the presence of
K2CO3 gives a mixture of the N-3 and the N-9 alkylated
isomers, with the former as the major product. The N-3
Scheme 1
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

November 2011
NMR and X-ray Structural Studies on 3-Benzyl-8-bromoadenine
1377
Figure 2. Adenine molecules connected by hydrogen bonds into one-dimensional chains or tapes. The indicated HꢂꢂꢂN distances are in the range
˚
2.09(4) ꢀ 2.16(3) A. It can be seen that molecule A participates in a larger number of strong interactions as the aromatic N atoms accept a total of
three H atoms compared with only one H atom for molecule B.
Melting points were determined with a Bu¨chi Melting Point B-
545 apparatus and are uncorrected. DMF was obtained from a
solvent purification system, MB SPS-800 from MBraun. Silica
gel for flash chromatography was purchased from Merck, Darm-
stadt, Germany (Merck No. 09385). All other reagents were
commercially available and used as received.
9-Benzyl-8-bromo-9H-purin-6-amine (2) and 3-benzyl-8-
bromo-3H-purin-6-amine (3). A mixture of 8-bromoadenine
[2] (220 mg, 1.03 mmol), DMF (5 mL) and potassium carbon-
ate (280 mg, 2.03 mmol) was stirred at ambient temperature
under N2 atmosphere for 30 min. Benzyl bromide (0.18 mL,
1.5 mmol) was added and the mixture stirred for another 4 h.
The mixture was filtered and the solvent evaporated in vacuo.
The residue was purified by column chromatography on silica
gel eluting with 0–5% MeOH in CH2Cl2 to give compounds 2
(45 mg, 15%) and 3 (128 mg, 42%).
selectivity is higher when the reaction is conducted at
ambient temperature compared with elevated tempera-
tures. Furthermore, it is shown that the 3-benzyl adenine
derivative exists as the amine tautomer both in the crys-
talline state as well as in solution (DMSO-d6), but with
restricted rotation around the N6AC6 bond.
EXPERIMENTAL
The 1H-NMR spectra were recorded at 300 MHz with a
Bruker Avance DPX 300 instrument and the decoupled 13C-
NMR spectra were recorded at 75 MHz using the instrument
mentioned above. The 1H-15N HSQC spectrum was recorded
with a Bruker AVII 600 instrument (pulse program hsqcetf3g-
psi, using sensitivity improvement). All NMR spectra were
obtained in DMSO-d6. Mass spectra under electron impact con-
ditions were recorded with a VG Prospec instrument at 70 eV
ionizing voltage, and are presented as m/z (% relative intensity).
9-Benzyl-8-bromo-9H-purin-6-amine (2). Colorless pow-
1
dery crystals, m.p. 238ꢁC (Lit. [4], 226–227ꢁC). H-NMR (300
MHz, DMSO-d6) d 8.16 (s, 1H, H-2), 7.47 (br s, 2H, NH2),
Figure 4. Resonance forms 3a0 and or 3a00 that may contribute to the
structure of tautomer 3a.
Figure 3. Possible tautomers of compound 3 in solution.
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet

¨
H.-S. M. Siah, C. H. Gorbitz, and L.-L. Gundersen
1378
Vol 48
7.33–7.21 (m, 5H, Ph), 5.35 (s, 2H, CH2); 13C-NMR (75 MHz,
DMSO-d6) d 154.7 (C-6), 153.0 (C-2), 150.9 (C-4), 135.9 (C-
8), 128.7 (Ph), 127.8 (Ph), 127.1 (Ph), 126.5 (Ph), 119.0 (C-5),
46.6 (CH2); ms: m/z 305/303 (26/26, Mþ), 91 (100).
3-Benzyl-8-bromo-3H-purin-6-amine (3). Colorless pow-
dery crystals, m.p. 233–235ꢁC (Lit. [4], 239–240.5ꢁC). 1H-
NMR (300 MHz, DMSO-d6) d 8.53 (s, 1H, H-2), 8.22 (br s,
1H, NH2), 8.10 (br s, 1H, NH2), 7.40–7.29 (m, 5H, Ph), 5.46
(s, 2H, CH2); 13C-NMR (75 MHz, DMSO-d6) d 153.6 (C-6),
149.8 (C-4), 144.0 (C-2), 139.3 (C-8), 135.8 (Ph), 128.7 (Ph),
128.1 (Ph), 127.8 (Ph), 121.5 (C-5), 52.0 (CH2); ms: m/z 305/
303 (28/28, Mþ), 91 (100).
X-ray crystallographic analysis for compound 3. Crystals
of 3 suitable for X-ray crystallography were obtained from a so-
lution of compound 3 in acetonitrile placed inside a larger vial
containing ethyl acetate. They were unstable at room temperature
due to loss of co-crystallized ethyl acetate solvent molecules, and
X-ray data collection with Apex-2 [11] was thus performed at
105 K. Apex II single crystal CCD-diffractometer, MoKa radia-
Acknowledgments. The Norwegian Research Council is greatly
acknowledged for financing of the Bruker Advance DPX 300
and the Bruker AVII 600 instruments used in this study. We
thank Prof. Frode Rise for obtaining the 1H-15N HSQC spectrum
and Matthew L. Read for assistance with crystallization of
compound 3.
REFERENCES AND NOTES
[1] Shaw, G. In Comprehensive Heterocyclic Chemistry; Potts,
K. T., Ed.; Pergamon Press: Oxford, 1984; Vol. 5, pp 499–605.
[2] Meszarosova, K.; Holy, A.; Masojidkova, M. Collect Czech
Chem Commun 2000, 65, 1109.
[3] Janeba, Z.; Holy, A.; Masojidkova, M. Collect Czech Chem
Commun 2000, 65, 1126.
[4] Havelkova, M.; Dvorak, D.; Hocek, M. Synthesis 2001,
1704.
[5] Li, X.; Vince, R. Bioorg Med Chem 2006, 14, 5742.
[6] Krouzelka, J.; Linhart, I.; Tobrman, T. J Heterocycl Chem
2008, 45, 789.
˚
tion (k ¼ 0.71069 A), 0.30 mm ꢃ 0.30 mm ꢃ 0.26 mm block-
shaped specimen, data integration and cell refinement with
SAINT-Plus [12], absorption correction by SADABS [13], struc-
ture solution by and least-squares refinement on F2 with
SHELXTL [14]. Solvent molecules are located on two different
inversion centers, each with a maximum allowed occupancy of
0.500 and form distinct channels running through the crystal
along the ab-diagonal. The geometries of independent solvent
molecules were constrained to be similar within a standard devia-
[7] Krouzelka, J.; Linhart, I. Heterocycles 2009, 78, 1205.
[8] Allen, F. H. Acta Crystallogr Sect B 2002, 58, 380.
[9] Glu¨senkamp, K.-H.; Kru¨ger, K.; Eberle, G.; Drosdziok, W.;
¨
¨
Jahde, E.; Grundel, O.; Neuhaus, A.; Boese, R.; Stellberg, P.; Rajew-
sky, M. F. Angew Chem Int Ed Engl 1993, 32, 1640.
[10] Bakkestuen, A. K.; Gundersen, L.-L.; Petersen, D.; Ute-
nova, B. T.; Vik, A. Org Biol Chem 2005, 3, 1025.
[11] Bruker AXS, Inc. APEX2; Bruker AXS, Inc.: Madison,
Wisconsin, USA, 2007.
˚
˚
tion of 0.002 A for bond lengths and 0.003 A for 1–3 distances.
3-Benzyl-8-bromo-3H-purin-6-amine ethyl acetate solvate:
C12H10BrN5ꢂ0.5C4H8O2, M ¼ 348.21, triclinic, P1, a ¼ 8.4937(6)
[12] Bruker AXS, Inc. SAINT-Plus; Bruker AXS, Inc.: Madison,
Wisconsin, USA, 2007.
ꢁ
˚
˚
˚
A, b ¼ 12.9096(9) A, c ¼ 15.2238(10) A, a ¼ 101.723(1) , b ¼
105.162(1)ꢁ, c ¼ 108.192(1)ꢁ, Z ¼ 4, Nobserved ¼ 5830, R[F2 >
2r(F2)] ¼ 0.058, wR(F2) ¼ 0.146, CCDC 786213.
[13] Bruker AXS, Inc. SADABS; Bruker AXS, Inc.: Madison,
Wisconsin, USA, 2007.
[14] Sheldrick, G. M. Acta Crystallogr Sect A 2008, 64, 112.
Journal of Heterocyclic Chemistry
DOI 10.1002/jhet